Efficient shapelet discovery for time series classification (extended abstract)

Li, Guozhong; Choi, Byron Koon Kau; Xu, Jianliang; Bhowmick, Sourav S; Chun, Kwok Pan; Wong, Grace LH

Abstract

Time-series shapelets are discriminative subsequences, recently found effective for time series classification (TSC). It is evident that the quality of shapelets is crucial to the accuracy of TSC. However, major research has focused on building accurate models from some shapelet candidates. To determine such candidates, existing studies are surprisingly simple, e.g., enumerating subsequences of some fixed lengths, or randomly selecting some subsequences as shapelet candidates. The major bulk of computation is then on building the model from the candidates. In this paper, we propose a novel efficient shapelet discovery method, called BSPCOVER, to discover a set of high-quality shapelet candidates for model building. We have conducted extensive experiments with well-known UCR time-series datasets and representative state-of-the-art methods. Results show that BSPCOVER speeds up the state-of-the-art methods by more than 70 times, and the accuracy is often comparable to or higher than existing works.

Citation

Li, G., Choi, B. K. K., Xu, J., Bhowmick, S. S., Chun, K. P., & Wong, G. L. (2021). Efficient shapelet discovery for time series classification (extended abstract). In 2021 IEEE 37th International Conference on Data Engineering (ICDE) (2336-2337). https://doi.org/10.1109/ICDE51399.2021.00254

• Conference Name: 2021 IEEE 37th International Conference on Data Engineering (ICDE)
• Start Date: Apr 19, 2021
• End Date: Apr 22, 2021
• Acceptance Date: Mar 19, 2020
• Online Publication Date: Jun 22, 2021
• Publication Date: Jun 22, 2021
• Deposit Date: Jan 19, 2022
• Publisher: Institute of Electrical and Electronics Engineers
• Volume: 2021-April
• Pages: 2336-2337
• Book Title: 2021 IEEE 37th International Conference on Data Engineering (ICDE)
• ISBN: 9781728191850
• DOI: https://doi.org/10.1109/ICDE51399.2021.00254
• Public URL: https://uwe-repository.worktribe.com/output/8545513
• Publisher URL: https://ieeexplore.ieee.org/abstract/document/9096567